The investigational drug brivanib is a dual inhibitor of vascular endothelial growth factor and fibroblast growth factor signaling, both implicated in hepatocellular carcinoma. The agent was recently evaluated in two phase III trials, one comparing first-line brivanib with sorafenib (Nexavar) in advanced hepatocellular carcinoma (BRISK-FL trial) and one comparing brivanib plus best supportive care vs placebo plus best supportive care in patients with advanced hepatocellular carcinoma in whom sorafenib therapy failed or was not tolerated (BRISK-PS trial).
In BRISK-FL, brivanib did not meet the criterion for noninferiority in overall survival compared with sorafenib. In BRISK-PS, brivanib did not improve overall survival, although benefits in response rate and time to progression were observed.
In the multinational phase III BRISK-FL trial, reported in the Journal of Clinical Oncology by Philip J.
Johnson, MD, of the Institute of Translational Medicine, University of Liverpool, and colleagues,1 1,155 patients with advanced hepatocellular carcinoma and no prior systemic therapy were randomly assigned to oral sorafenib at 400 mg twice daily (n = 578) or oral brivanib at 800 mg once daily (n = 577). The primary endpoint was overall survival.
In total, 62% of patients were from Asia, 23% from Europe, 13% from the Americas, 0.8% from Australia, and 0.6% from Africa. The brivanib and sorafenib groups were generally well matched for age (median, 61 and 60 years), sex (84% male in both), region (eg, Asia for 60% and 64%), Eastern Cooperative Oncology Group (ECOG) performance status (0 in 64% and 61%), Barcelona Clinical Liver Cancer stage (eg, C in 77% and 78%), Child-Pugh class (A in 92% in both), macrovascular invasion (no in 73% of both), portal vein invasion/thrombosis (yes in 19% of both), distant metastasis (49% and 50%), regional lymph node metastasis (27% and 28%), extrahepatic spread/macrovascular invasion (63% and 62%), alpha-fetoprotein level (median, 142 and 180 ng/mL), risk factors (any in 78% and 75%, hepatitis B virus in 44% and 45%), and previous nonsystemic treatment (55% and 56%; liver resection in 28% and 30%, transcatheter arterial chemoembolization in 35% and 36%).
A total of 1,150 patients were treated in the BRISK-FL trial (per-protocol population). At the time of the final analysis, 6% of the brivanib group and 11% of the sorafenib group remained on study. The most common reasons for study discontinuation were disease progression (46% in the brivanib group and 53% in the sorafenib group) and study drug toxicity (24% and 15%).
Median duration of treatment was 3.2 months in the brivanib group and 4.1 months in the sorafenib group. Poststudy systemic treatments were received by 22% of the brivanib group and 21% of the sorafenib group, and nonsystemic treatments were received by 19% and 17%.
Overall Survival Noninferiority
The primary endpoint of overall survival noninferiority for brivanib vs sorafenib in the per-protocol population was not met (hazard ratio [HR] = 1.06, 95.8% confidence interval [CI] = 0.93–1.22, P = .373), based on the prespecified noninferiority margin (upper CI limit for HR ≤ 1.08). Median overall survival was 9.5 months in the brivanib group and 9.9 months in the sorafenib group. Overall survival results were similar in the intent-to-treat population (HR = 1.07, 95.8% CI = 0.94–1.23, P = .312). Results for subgroups were consistent with those for the overall study population.
On multivariate analysis, significant prognostic factors for overall survival were alpha-fetoprotein, tumor morphologic features, size of the largest nodule, Child-Pugh score, and major portal vein invasion. After adjusting for baseline factors, the effect of brivanib vs sorafenib remained unchanged (HR = 1.09, 95% CI = 0.95–1.25).
Median time to progression was 4.2 months in the brivanib group and 4.1 months in the sorafenib group (P = .853), objective response rate was 12% vs 9% (P = .057), and disease control rate was 66% vs 65% (P = .874). Among patients with baseline alpha-fetoprotein ≥ 200 ng/mL and at least one on-study alpha-fetoprotein assessment, alpha-fetoprotein reduction ≥ 50% from baseline was observed in 58% of brivanib patients vs 31% of sorafenib patients. Similar reductions were observed when the baseline cutoff used was the upper limit of normal or 400 ng/mL.
Among adverse events of any grade, hand-foot skin reaction, alopecia, rash, and pyrexia were more frequent in sorafenib patients and decreased appetite, fatigue, hypertension, nausea, vomiting, hyponatremia, headache, dysphonia, and dizziness were more frequent in brivanib patients. Diarrhea, abdominal pain, constipation, hyperbilirubinemia, elevated aspartate aminotransferase (AST), elevated alanine aminotransferase (ALT), and weight loss occurred at a similar rate in the two groups. The most frequent grade 3 or 4 adverse events were hyponatremia (23% in brivanib group vs 9% in sorafenib group), AST elevation (14% vs 17%), fatigue (15% vs 7%), hand-foot skin reaction (2% vs 15%), and hypertension (13% vs 5%).
Treatment was discontinued due to adverse events in 43% of brivanib patients and 33% of sorafenib patients, with the most common reasons being fatigue (5%), hyponatremia (2%), decreased appetite (2%), hyperbilirubinemia (2%), and AST elevations (2%) in the brivanib group and hyperbilirubinemia (3%) and AST elevations (2%) in the sorafenib group. Dose reduction occurred in 49% and 50% of patients, and dose interruption occurred in 58% of both groups.
The investigators concluded: “This study did not meet its primary [overall survival] objective in the first-line treatment of advanced [hepatocellular carcinoma], based on a noninferiority statistical design, but it did show similar antitumor activity for brivanib and sorafenib, based on time to progression, objective response rate, and disease control rate. Brivanib had an acceptable safety profile; however, it was less well-tolerated than sorafenib.”
This multinational phase III trial—the first phase III brivanib trial in patients who have already received sorafenib—enrolled 395 patients with advanced hepatocellular carcinoma whose disease progressed on/after sorafenib or who were intolerant of sorafenib. Patients were randomly assigned (2:1) to receive oral brivanib at 800 mg once daily plus best supportive care (n = 263) or placebo plus best supportive care (n = 132). The primary endpoint was overall survival. The study was reported in the Journal of Clinical Oncology by Josep M. Llovet, MD, of the Icahn School of Medicine at Mount Sinai in New York, and colleagues.2
Overall, 42% of patients were from Europe, 41% from Asia, and 17% from the Americas. The brivanib and placebo groups were generally well matched for age (median, 64 and 62 years), sex (82% and 86% male), race (46% and 50% white, 48% and 45% Asian), reason for sorafenib discontinuation (progression in 86% and 88%), prior nonsystemic treatment (any in 80% and 73%; transcatheter chemoembolization in 57% and 49%, liver resection in 39% and 36%), ECOG performance status (0 in 57% and 61%), Barcelona Clinical Liver Cancer stage (C in 87% and 85%), Child-Pugh class (A in 92% and 91%), distant metastasis (65% and 64%), regional lymph node metastasis (35% and 36%), and risk factors (any in 83% and 85%, hepatitis B virus infection in 39% and 34%). The brivanib group had more patients with vascular invasion (31% vs 18%), including portal vein invasion/ thrombosis (25% vs 12%), and a higher alpha-fetoprotein level (median, 204 vs 100 ng/mL).
Study treatment was discontinued in 90% of brivanib patients and 92% of placebo patients, with the most common reasons being disease progression (50% and 70%) and study drug toxicity (23% and 7%). Median treatment duration was 3.1 months for brivanib and 2.5 months for placebo. After discontinuation of study treatment, 27% of the brivanib group and 35% of the placebo group received systemic therapies and 18% and 20% received nonsystemic treatments.
Overall Survival Outcome
Overall survival was not significantly improved in the brivanib vs placebo group (HR = 0.89, 95.8% CI = 0.69–1.15, P = .331); median overall survival was 9.4 and 8.2 months, respectively. Overall survival results across prespecified subgroups were generally consistent with the primary overall survival analysis.
On multivariate analysis, alpha-fetoprotein (< 200 ng/mL vs ≥ 200 g/mL, P < .001) and portal vein invasion (no vs yes, P < .001) were significant prognostic factors for overall survival. After adjustment for all prespecified factors, the hazard ratio for overall survival for brivanib vs placebo was 0.81 (P = .1044). Post hoc analyses using poststudy treatments as a time-dependent covariate did not show an impact of poststudy treatment on overall survival.
Time to progression was significantly longer in the brivanib group (median, 4.2 vs 2.7 months, HR = 0.56, P < .001), including after adjustment for baseline prognostic factors (HR = 0.56, 95% CI = 0.42–0.76).
The objective response rate was also significantly higher with brivanib treatment (10% vs 2%, odds ratio [OR] = 5.75, P = .0030), as was disease control rate (61% vs 40%, OR = 2.38, P < .001).
Among patients with baseline alpha-fetoprotein greater than the upper limit of normal, a reduction of ≥ 50% from baseline occurred in 54% of brivanib patients and 7% of placebo patients.
Study discontinuation due to treatment-related adverse events occurred in 23% of brivanib patients and 7% of placebo patients. The most frequent treatment-related grade 3 or 4 adverse events in the brivanib group included hypertension (17% vs 2%), fatigue (13% vs 1%), hyponatremia (11% vs 2%), and decreased appetite (10% vs 2%). The most frequent grade 3 or 4 laboratory abnormalities were hyponatremia (27% vs 14%), AST increase (27% vs 24%), and hyperbilirubinemia (21% vs 18%).
Six deaths on study were considered by investigators as possibly related to treatment, all of which were in the brivanib arm and occurred within 30 days of the final dose. Two of the deaths were a result of encephalopathy, and one each was due to liver failure, acidosis, sudden death, and coma/cerebral edema.
The investigators concluded: “In patients with [hepatocellular carcinoma] who had been treated with sorafenib, brivanib did not significantly improve [overall survival]. The observed benefit in the secondary outcomes of [time to progression] and objective response rate warrants further investigation…. The results of our trial may inform the design of future studies in this patient population.” ■
Disclosure: The studies were supported by Bristol-Myers Squibb. For full disclosures of the study authors, visit jco.ascopubs.org.
1. Johnson PJ, Qin S, Park J-W, et al: Brivanib versus sorafenib as first-line therapy in patients with unresectable, advanced hepatocellular carcinoma: Results from the randomized phase III BRISK-FL study. J Clin Oncol 31:3517-3524, 2013.
2. Llovet JM, Decaens T, Raoul J-L, et al: Brivanib in patients with advanced hepatocellular carcinoma who were intolerant to sorafenib or for whom sorafenib failed: Results from the randomized phase III BRISK-PS study. J Clin Oncol 31:3509-3516 2013.